Computers, fax machines, printers and other electronic devices are routinely connected by communications cables to network equipment and/or to external networks such as the Internet. FIG. 1 illustrates the manner in which a computer 10 may be connected to network equipment 20 using conventional communications plug/jack connections. As shown in FIG. 1, the computer 10 is connected by a patch cord assembly 11 to a communications jack 30 that is mounted in a wall plate 19. The patch cord assembly 11 comprises a communications cable 12 that contains a plurality of individual conductors (e.g., insulated copper wires) and two communications plugs 13, 14 that are attached to the respective ends of the cable 12. The communications plug 13 is inserted into a communications jack (not pictured in FIG. 1) that is provided in the computer 10, and the communications plug 14 inserts into a plug aperture 32 in the front side of the communications jack 30. The plug contacts (which are commonly referred to as “blades”) of communications plug 14 (which are exposed through the slots 15 on the top and front surfaces of communications plug 14) mate with respective contacts (not visible in FIG. 1) of the communications jack 30 when the communications plug 14 is inserted into the plug aperture 32. The blades of communications plug 13 similarly mate with respective contacts of the communications jack (not pictured in FIG. 1) that is provided in the computer 10.
The communications jack 30 includes a back-end connection assembly 50 that receives and holds conductors from a cable 60. As shown in FIG. 1, each conductor of cable 60 is individually pressed into a respective one of a plurality of slots provided in the back-end connection assembly 50 to establish mechanical and electrical connection between each conductor of cable 60 and the communications jack 30. The other end of each conductor in cable 60 may be connected to, for example, the network equipment 20. The wall plate 19 is typically mounted on a wall (not shown) of a room or office of, for example, an office building, and the cable 60 typically runs through conduits in the walls and/or ceilings of the building to a room in which the network equipment 20 is located. The patch cord assembly 11, the communications jack 30 and the cable 60 provide a plurality of signal transmission paths over which information signals may be communicated between the computer 10 and the network equipment 20. It will be appreciated that typically one or more patch panels or switches, along with additional communications cabling, would be included in the electrical path between the cable 60 and the network equipment 20. However, for ease of description, these additional elements have been omitted from FIG. 1 and the cable 60 is instead shown as being directly connected to the network equipment 20.
In many electrical communications systems that are used to interconnect computers, network equipment, printers and the like, the information signals are transmitted between devices over a pair of conductors (hereinafter a “differential pair” or simply a “pair”) rather than over a single conductor. The signals transmitted on each conductor of the differential pair have equal magnitudes, but opposite phases, and the information signal is embedded as the voltage difference between the signals carried on the two conductors of the pair. When signals are transmitted over a conductor (e.g., an insulated copper wire) in a communications cable, electrical noise from external sources such as lightning, electronic equipment, radio stations, etc. may be picked up by the conductor, degrading the quality of the signal carried by the conductor. When the signal is transmitted over a differential pair of conductors, each conductor in the differential pair often picks up approximately the same amount of noise from these external sources. Because approximately an equal amount of noise is added to the signals carried by both conductors of the differential pair, the information signal is typically not disturbed, as the information signal is extracted by taking the difference of the signals carried on the two conductors of the differential pair; thus, the noise signal is cancelled out by the subtraction process.
The cables and connectors in many, if not most, high speed communications systems include eight conductors that are arranged as four differential pairs. Channels are formed by cascading plugs, jacks and cable segments to provide connectivity between two end devices. In these channels, when a plug mates with a jack, the proximities and routings of the conductors and contacting structures within the jack and/or plug can produce capacitive and/or inductive couplings. Moreover, since four differential pairs are usually bundled together in a single cable, additional capacitive and/or inductive coupling may occur between the differential pairs within each cable. These capacitive and inductive couplings in the connectors and cabling give rise to another type of noise that is called “crosstalk.”
“Crosstalk” in a communication system refers to unwanted signal energy that is induced onto the conductors of a first “victim” differential pair from a signal that is transmitted over a second “disturbing” differential pair. The induced crosstalk may include both near-end crosstalk (NEXT), which is the crosstalk measured at an input location corresponding to a source at the same location (i.e., crosstalk whose induced voltage signal travels in an opposite direction to that of an originating, disturbing signal in a different path), and far-end crosstalk (FEXT), which is the crosstalk measured at the output location corresponding to a source at the input location (i.e., crosstalk whose signal travels in the same direction as the disturbing signal in the different path). Both types of crosstalk comprise an undesirable noise signal that interferes with the information signal on the victim differential pair.
A variety of techniques may be used to reduce crosstalk in communications systems such as, for example, tightly twisting the paired conductors in a cable, whereby different pairs are twisted at different rates that are not harmonically related, so that each conductor in the cable picks up approximately equal amounts of signal energy from the two conductors of each of the other differential pairs included in the cable. If this condition can be maintained, then the crosstalk noise may be significantly reduced, as the conductors of each differential pair carry equal magnitude, but opposite phase signals such that the crosstalk added by the two conductors of a differential pair onto the other conductors in the cable tends to cancel out.
While such twisting of the conductors and/or various other known techniques may substantially reduce crosstalk in cables, most communications systems include both cables and communications connectors (i.e., jacks, plugs and connecting blocks, etc.) that interconnect the cables and/or connect the cables to computer hardware. Unfortunately, the connector configurations that were adopted years ago generally did not maintain the conductors of each differential pair a uniform distance from the conductors of the other differential pairs in the connector hardware. Moreover, in order to maintain backward compatibility with connector hardware that is already installed, the connector configurations have, for the most part, not been changed. As such, the conductors of each differential pair tend to induce unequal amounts of crosstalk on each of the other conductor pairs in current and pre-existing connectors. As a result, many current connector designs generally introduce some amount of NEXT and FEXT crosstalk.
Pursuant to certain industry standards (e.g., the TIA/EIA-568-B.2-1 standard approved Jun. 20, 2002 by the Telecommunications Industry Association), each jack, plug and cable segment in a communications system may include a total of eight conductors 1-8 that comprise four differential pairs. The industry standards specify that, in at least the connection region where the contacts (blades) of a modular plug mate with the contacts of the modular jack (referred to herein as the “plug jack mating region”), the eight conductors are aligned in a row, with the four differential pairs specified as depicted in FIG. 2. As known to those of skill in the art, under the TIA/EIA 568 type B configuration, conductors 4 and 5 in FIG. 2 comprise pair 1, conductors 1 and 2 comprise pair 2, conductors 3 and 6 comprise pair 3, and conductors 7 and 8 comprise pair 4. As known to those of skill in the art, conductors 1, 3, 5 and 7 comprise “tip” conductors, and conductors 2, 4, 6 and 8 comprise “ring” conductors.
As shown in FIG. 2, in the plug-jack mating region, the conductors of the differential pairs are not equidistant from the conductors of the other differential pairs. By way of example, conductors 1 and 2 of pair 2 are different distances from conductor 3 of pair 3. Consequently, differential capacitive and/or inductive coupling occurs between the conductors of pairs 2 and 3 that generate both NEXT and FEXT. Similar differential coupling occurs with respect to the other differential pairs in the modular plug and the modular jack. This differential coupling typically occurs in the blades of the modular plugs and in at least a portion of the contacts of the modular jack.
As the operating frequencies of communications systems increased, crosstalk in the plug and jack connectors became a more significant problem. To address this problem, communications jacks were developed that included compensating crosstalk circuits that introduced compensating crosstalk that was used to cancel much of the “offending” crosstalk that was being introduced in the plug-jack mating region. In particular, in order to cancel the “offending” crosstalk that is generated in a plug-jack connector because a first conductor of a first differential pair inductively and/or capacitively couples more heavily with a first of the two conductors of a second differential pair than does the second conductor of the first differential pair, jacks were designed so that the second conductor of the first differential pair would capacitively and/or inductively couple with the first of the two conductors of the second differential pair later in the jack to provide a “compensating” crosstalk signal. As the first and second conductors of the differential pair carry equal magnitude, but opposite phase signals, so long as the magnitude of the “compensating” crosstalk signal that is induced in such a fashion is equal to the magnitude of the “offending” crosstalk signal, then the compensating crosstalk signal that is introduced later in the jack may substantially cancel out the offending crosstalk signal.
FIG. 3 is a schematic diagram of a plug-jack connector 60 (i.e., an RJ-45 communications plug 70 that is mated with an RJ-45 communications jack 80) that illustrate how the above-described crosstalk compensation scheme may work. As shown by the arrow in FIG. 3 (which represents the time axis for a signal flowing from the plug 70 to the jack 80), crosstalk having a first polarity (here arbitrarily shown by the “+” sign as having a positive polarity) is induced from the conductor(s) of a first differential pair onto the conductor(s) of a second differential pair. By way of example, when a signal is transmitted on pair 3 of plug 70, in both the plug 70 and in the plug-jack mating region portion of the jack 80, the signal on conductor 3 of pair 3 will induce a larger amount of current onto conductor 4 of pair 1 than conductor 6 of pair 3 will induce onto conductor 4 of pair 1, thereby resulting in an “offending” crosstalk signal on pair 1. By arranging the conductive paths in a later part of the jack 80 to include a capacitor between, for example, conductors 3 and 5 and/or to have inductive coupling between conductors 3 and 5, it is possible to introduce one or more “compensating” crosstalk signals in the jack 80 that will at least partially cancel the offending crosstalk signal on pair 1. An alternative method for generating such a compensating crosstalk signal would be to design the jack 80 to provide capacitive and/or inductive coupling between conductors 4 and 6, as the signal carried by conductor 6 has a polarity that is opposite the signal carried by conductor 3.
While the simplified example of FIG. 3 discusses methods of providing compensating crosstalk that cancels out the differential crosstalk induced from conductor 3 to conductor 4 (i.e., part of the pair 3 to pair 1 crosstalk), it will be appreciated that the industry standardized connector configurations result in offending crosstalk between various of the differential pairs, and compensating crosstalk circuits are typically provided in the jack for reducing the offending crosstalk between more than one pair combination.
FIG. 4 is a schematic graph that illustrates the offending crosstalk signal and the compensating crosstalk signal that are discussed above with respect to FIG. 3 as a function of time. In the plug blades and in the plug-jack mating region of the jack, the offending crosstalk signal that is discussed in the example above is the signal energy induced from conductor 3 onto conductor 4 minus the signal energy induced from conductor 6 onto conductor 4. This offending crosstalk is represented by vector A0 in FIG. 4, where the length of the vector represents the magnitude of the crosstalk and the direction of the vector (up or down) represents the polarity (positive or negative) of the crosstalk. It will be appreciated that the offending crosstalk will typically be distributed to some extent over the time axis, as the differential coupling typically starts at the point where the wires of the cable (e.g., conductors 3-6) are untwisted and continues through the plug blades and into the jack contact region of the jack 80 (and perhaps even further into the jack 80). However, for ease of description, this distributed crosstalk is represented as a single crosstalk vector A0 having a magnitude equal to the sum of the distributed crosstalk that is located at the weighted midpoint of the differential coupling region (referred to herein as a “lumped approximation”).
As is further shown in FIG. 4, the compensating crosstalk circuit in the jack 80 (e.g., a capacitor between conductors 4 and 6) induces a second crosstalk signal onto pair 1 which is represented by the vector A1 in FIG. 4. As the crosstalk compensation circuit is located after the jackwire contacts (with respect to a signal travelling in the forward direction from the plug 70 to the jack 80), the compensating crosstalk vector A1 is located farther to the right on the time axis. The compensating crosstalk vector A1 has a polarity that is opposite to the polarity of the offending crosstalk vector A0 as conductors 3 and 6 carry opposite phase signals.
The signals carried on the conductors are alternating current signals, and hence the phase of the signal changes with time. As the compensating crosstalk circuit is typically located quite close to the plug-jack mating region (e.g., less than an inch away), the time difference (delay) between the offending crosstalk region and the compensating crosstalk circuit is quite small, and hence the change in phase likewise is small for low frequency signals. As such, the compensating crosstalk signal can be designed to almost exactly cancel out the offending crosstalk with respect to low frequency signals (e.g., signals having a frequency less than 100 MHz).
However, for higher frequency signals, the phase change between vectors A0 and A1 can become significant. Moreover, in order to meet the increasing throughput requirements of modern computer systems, there is an ever increasing demand for higher frequency connections. FIG. 5A is a vector diagram that illustrates how the phase of compensating crosstalk vector A1 will change by an angle φ due to the time delay between vectors A0 and A1. As a result of this phase change φ, vector A1 is no longer offset from vector A0 by 180°, but instead is offset by 180°-φ. Consequently, compensating crosstalk vector A1 will not completely cancel the offending crosstalk vector A0. This can be seen graphically in FIG. 5B, which illustrates how the addition of vectors A0 and A1 still leaves a residual crosstalk vector. FIG. 5B also makes clear that the degree of cancellation decreases as cp gets larger. Thus, due to the increased phase change at higher frequencies, the above-described crosstalk compensation scheme cannot fully compensate for the offending crosstalk.
U.S. Pat. No. 5,997,358 to Adriaenssens et al. (hereinafter “the '358 patent”) describes multi-stage crosstalk compensating schemes for plug-jack connectors that can be used to provide significantly improved crosstalk cancellation, particularly at higher frequencies. The entire contents of the '358 patent are hereby incorporated herein by reference as if set forth fully herein. Pursuant to the teachings of the '358 patent, two or more stages of compensating crosstalk are added, usually in the jack, that together reduce or substantially cancel the offending crosstalk at the frequencies of interest. The compensating crosstalk can be designed, for example, into the lead frame wires of the jack and/or into a printed wiring board that is electrically connected to the lead frame.
As discussed in the '358 patent, the magnitude and phase of the compensating crosstalk signal(s) induced by each stage are selected so that, when combined with the compensating crosstalk signals from the other stages, they provide a composite compensating crosstalk signal that substantially cancels the offending crosstalk signal over a frequency range of interest. In embodiments of these multi-stage compensation schemes, the first compensating crosstalk stage (which can include multiple sub-stages) has a polarity that is opposite the polarity of the offending crosstalk, while the second compensating crosstalk stage has a polarity that is the same as the polarity of the offending crosstalk.
FIG. 6A is a schematic graph of crosstalk versus time that illustrates the location of the offending and compensating crosstalk (depicted as lumped approximations) if the jack of FIG. 3 is modified to implement multi-stage compensation. As shown in FIG. 6A, the offending crosstalk signal that is induced in the plug and in the plug-jack mating region can be represented by the vector B0 which has a magnitude equal to the sum of the distributed offending crosstalk and which is located at the weighted midpoint of the coupling region where the offending crosstalk is induced. As is further shown in FIG. 6A, the compensating crosstalk circuit in the jack induces a second crosstalk signal which is represented by the vector B1. As the crosstalk compensation circuit is located after the jackwire contacts (with respect to a signal travelling in the forward direction), the compensating crosstalk vector B1 is located farther to the right on the time axis. The compensating crosstalk vector B1 has a polarity that is opposite to the polarity of the offending crosstalk vector B0. Moreover, the magnitude of the compensating crosstalk vector B1 is larger than the magnitude of the offending crosstalk vector B0. Finally, a second compensating crosstalk vector B2 is provided that is located even farther to the right on the time axis. The compensating crosstalk vector B2 has a polarity that is opposite the polarity of crosstalk vector B1, and hence that is the same as the polarity of the offending crosstalk vector B0.
FIG. 6B is a vector summation diagram that illustrates how the multi-stage compensation crosstalk vectors B1 and B2 of FIG. 6A can cancel the offending crosstalk vector B0 at a selected frequency. FIG. 6B takes the crosstalk vectors from FIG. 6A and plots them on a vector diagram that visually illustrates the magnitude and phase of each crosstalk vector. In FIG. 6B, the dotted line versions of vectors B1 and B2 are provided to show how the three vectors B0, B1 and B2 may be designed to sum to approximately zero at a selected frequency. In particular, as shown in FIG. 6B, the first compensating crosstalk stage (B1) significantly overcompensates the offending crosstalk. The second compensating crosstalk stage (B2) is then used to bring the sum of the crosstalk back to the origin of the graph (indicating substantially complete cancellation at the selected frequency). The multi-stage (i.e., two or more) compensation schemes disclosed in the '358 patent thus can be more efficient at reducing the NEXT than schemes in which the compensation is added at a single stage.
The first compensating stage can be placed in a variety of locations. U.S. Pat. Nos. 6,350,158; 6,165,023; 6,139,371; 6,443,777 and 6,409,547 disclose communications jacks having crosstalk compensation circuits implemented on or connected to the free ends of the jackwire contacts, The '358 patent discloses communications jacks having crosstalk compensation circuits implemented on a printed circuit board that are connected to the mounted ends of the jackwire contacts.